This invention pertains to absorption dynamometers, and in particular, to torque measuring devices used in combination with air absorption dynamometers.
Means of measuring torque generally fall into two categories: in-line (rotating) and reaction (stationary). Reaction torque measurements are used with air dynamometers, but they are inaccurate, being subject to errors introduced by aerodynamic effects associated with the exhaust of the engine being tested, the exhaust of the dynamometer, and local external air velocities. In-line torque measurements have also been used, but prior in-line torque meters have various drawbacks.
In one type of in-line torque meter, strain gages are applied to a rotating member through which the torque is transmitted. The strain gages sense the twist of this member by minute changes in resistance which are measured as changes in voltage in a bridge circuit. Because of the high rotating speed in the dynamometer, communicating the strain gage signal from the rotor to the stator is difficult and the available mechanisms available (slip rings, telemetry) are sources of unreliability and error.
In another type of in-line torque meter, two concentric shafts are located between the engine and the dynamometer. The inner shaft is connected to both the engine and the dynamometer. It transmits the torque and in so doing is twisted in proportion to the torque. The outer shaft is connected at only one end and transmits no torque. The angular displacement between these concentric shafts is a measure of the applied torque. This angular displacement is measured by comparing the phase of two sensors located on adjacent stationary structure. One of the sensors observes the passing of teeth on the inner shaft, the other the teeth on the outer shaft. As these teeth are angularly displaced relative to each other in proportion to the applied torque, the phase relationship of the signals from the two sensors is also displaced in proportion to the torque.
This type of in-line torque meter suffers from two drawbacks. The use of two sensors introduces an inherent error into the measurement. As the quantity being measured is the difference between the signals from the two sensors, any errors due to drift in the sensors themselves are combined into the measurement error.
The foregoing illustrates limitations known to exist in present dynamometer torque measurement systems. Thus, it is apparent that it would be advantageous to provide an alternative directed to overcoming one or more of the limitations set forth above. Accordingly, a suitable alternative is provided including features more fully disclosed hereinafter.